Fuzzy Dynamical Modeling Techniques for Nonlinear Control Systems and Their Applications to Multiple-Input, Multiple-Output (Mimo) Systems

This chapter proposes a new approach to fuzzy adaptive controller design using only system input–output data. The design procedure consists of three steps: First, a fuzzy ARMAX model is identified using the available data; then, a fuzzy controller is derived based on a combination of sliding mode control (SMC) theory and fuzzy control methodology; finally, the controller is equipped with an adaptive mechanism to update all control parameters online. The resulting control systems generally possess strong model-adaptation capability, good tracking control performance, and robustness against perturbations and uncertainties. This approach is applicable to general multiple-input, multiple-output (MIMO) nonlinear and uncertain systems. In this chapter, the basic problem of modeling and control of an unknown multivariable dynamic system is studied with analysis, design, and simulations. The new controller only used input-output data, but has a strong prediction ability and satisfactory tracking performance. Several simulation examples have been shown to demonstrate the merits of the proposed FARMAX (fuzzy ARMAX).